What is the best local model for coding on the Instinct Mi60?

On the Instinct Mi60, Qwen3.6-35B-A3B-UD-Q5_K_XL ranks first for coding in our benchmarks (71.8/100 at ~58.4 tok/s).

What is the best local model for agentic workflows on the Instinct Mi60?

On the Instinct Mi60, Qwen3.6-35B-A3B-UD-Q5_K_XL ranks first for agentic workflows in our benchmarks (78.8/100 at ~58.4 tok/s).

Hardware Benchmarks

Compare Instinct Mi60 benchmark results and see which models actually earn the best completed quality scores.

GPU-specific specifications for local LLM planning.

Radeon Instinct MI60

Models we recommend for Instinct Mi60.

Fastest

26B

76.8 tok/s

Highest token generation speed at 76.8 tok/s.

Best quality

35B

58.4 tok/sQuality 76.7

Highest completed quality score at 76.7.

Which models actually run best on the Instinct Mi60, by task, from community benchmark data.

What is the best local model for coding on the Instinct Mi60?: On the Instinct Mi60, Qwen3.6-35B-A3B-UD-Q5_K_XL ranks first for coding in our benchmarks (71.8/100 at ~58.4 tok/s).
What is the best local model for agentic workflows on the Instinct Mi60?: On the Instinct Mi60, Qwen3.6-35B-A3B-UD-Q5_K_XL ranks first for agentic workflows in our benchmarks (78.8/100 at ~58.4 tok/s).

Completed quality-ranked benchmark results for Radeon Instinct MI60.